Anthropic Announces the Lightweight, Cost-Effective 'Claude Haiku 4.5,' Delivering Performance on a par with Claude Sonnet 4 at One-Third the Cost and More Than Twice the Speed



On October 16, 2025 local time, AI development company Anthropic announced ' Claude Haiku 4.5 .' It achieves the same coding performance as

Claude Sonnet 4 , the most advanced AI model as of May 2025, at just one-third the cost and more than twice the speed.

Introducing Claude Haiku 4.5 \ Anthropic
https://www.anthropic.com/news/claude-haiku-4-5



Introducing Claude Haiku 4.5 - YouTube


Anthropic's Claude Haiku 4.5 matches May's frontier model at fraction of cost - Ars Technica
https://arstechnica.com/ai/2025/10/anthropics-claude-haiku-4-5-matches-mays-frontier-model-at-fraction-of-cost/

Claude, Anthropic's AI model, has three models: the performance-oriented 'Opus,' the balanced 'Sonnet' that combines performance and cost, and the lightest and most cost-conscious 'Haiku.' The cost-conscious lightweight model Haiku now has its cutting-edge model, Claude Haiku 4.5.

Claude Haiku 4.5 outperforms Claude Sonnet 4 in certain tasks, such as computer use, making applications like Claude for Chrome faster and more convenient than ever.



'Users who leverage AI for real-time, low-latency tasks like chat assistants, customer service agents, and pair programming will appreciate the combination of high intelligence and incredible speed in Claude Haiku 4.5,' Anthropic said. 'Claude Code users will find that Claude Haiku 4.5 delivers a significantly more responsive coding experience for everything from multi-agent projects to rapid prototyping.'

The graph below compares the accuracy (vertical axis) of multiple AI models in software engineering coding tasks. Claude Haiku 4.5 (orange) generated code with an accuracy of 73.3%, lower than

Claude Sonnet 4.5 (77.2%), but outperformed Claude Sonnet 4 (72.7%). It also outperformed competitors' AI models, such as GPT-5 and Gemini 2.5 Pro .



Below are the results of comparing Claude Haiku 4.5 with competing AI models based on various benchmark tests. Although it does not reach the top model, Claude Sonnet 4.5, it shows that it performs equal to or better than Claude Sonnet 4 in most benchmark tests.



The Claude Sonnet 4.5, released at the end of September 2025, is the most advanced model from Anthropic at the time of writing. Meanwhile, the newly released Claude Haiku 4.5 offers a new option for users looking for cutting-edge performance with greater cost efficiency.

Anthropic is also developing new ways to connect multiple AI models. For example, Claude Sonnet 4.5 can break down complex problems into multi-step plans, and then teams of multiple Claude Haiku 4.5s can be organized to complete subtasks in parallel.

Claude Haiku 4.5 will be available from October 16th and developers can access it via the Claude API as 'claude-haiku-4-5.' The fee is $1 per million input tokens and $5 per million output tokens.

in Video,   Software, Posted by logu_ii